Minimum important difference is minimally important in sample size calculations.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Research Methodology

Background:

• Sample size calculations are crucial for randomized controlled trials (RCTs).
• A common practice is using the minimum important difference (MID) to set the assumed benefit for sample size calculations.
• This approach is widely believed to be correct but is often misguided.

Purpose of the Study:

• To clarify the appropriate use of the minimum important difference (MID) in sample size calculations for RCTs.
• To emphasize the importance of realistic benefit estimates in determining sample size.
• To guide researchers in conducting valid and powerful clinical trials.

Main Methods:

• The study critically evaluates the conventional method of using MID for assumed benefit in sample size calculations.
• It contrasts this with the necessity of employing realistic benefit estimates.
• The paper discusses the impact of assumed benefit on the true power of an RCT.

Main Results:

• The MID should only inform the decision of whether to conduct a trial at all (i.e., if the sample size should be zero).
• The assumed benefit in sample size calculations should reflect a realistic estimate of the true benefit.
• Using realistic benefit estimates ensures the calculated sample size yields a true power close to the target power.

Conclusions:

• Misguided use of MID in sample size calculations leads to inaccurate true power.
• Researchers should prioritize realistic benefit estimation over MID for valid sample size calculations.
• RCTs should only proceed if a realistic benefit estimate suggests a meaningful outcome is likely.